Light mod Denon 3910..

[Ryssen]

I have to change the laser on my 3910,while the hood is off I might at well install the XO3 that I have laying around.And change some regulators to the LM317/337 type.But firts some questions,the supply to the XO3,I could take it from the +12v line,but the sonic result will be better with a separate trafo and rectifier,or?
I se that the opams in the analog output runs on + - 9v,those regulators can be upgraded.
Then I assume the digital part runns on +5V,but there are 2 different 5v supplys(Green on the schema,+ - 9v, red)which one of the 5v regulators to uppgrade,or both?



XO3 clock in on XTI,remove C112,C118 and R127,right?

[Ryssen]

Anyone tried this:
it even got better ripple rejection than TL317!TL780-05

[anatech]

Hi Ryssen,
I think you should look at the HF performance of the regulators. Bypassing the electrolytic caps might make a far better improvement than the basic regulator. This means leaving the regulators stock and bypassing may be better all around. Just don't get silly with the caps. Small box polyesters will be just fine.

I haven't looked at the regulator specs, but IC regulators generally have trouble at the higher frequencies.

-Chris

[Ryssen]

Yes,the ripple current is messaured at 120hz..And the only one of 7805,780-05 and LT317 that has a diagram of the ripple at different frekvensys is LT317,and it shows that the ripple rejection is down about 20dB at 20Khz compared to 120hz..Then I wonder how bad the 7805 that is the one in the player.
I have some PCB´s for a discrete regulator,maybe use them?
And I`am going to e-mail texas instruments for some more data on the
LM780-05.
Yes,I am going to bypass the electrolytic caps,thats my favorit.

[anatech]

Hi Ryssen,
I don't know what the high frequency behavior is on these. Adding a choke and cap after these might be a good idea, and before. That reduces the importance of regulator performance at higher frequencies.

I have some PCB´s for a discrete regulator,maybe use them?

Possibly. There is a chance the performance might be worse than what is in there. Depends. As long as you don't cause damage to the PCB, go ahead and try them. Should be interesting to see what your results are. If you can, try to look at the supplies before and after.

-Chris

[Hartono]

put snubber on the power supply diode, or better change to fast "soft"recovery type. impotant to notice the "soft" when you decided to buy one.

Hartono

[Ryssen]

Adding a choke and cap after these might be a good idea, and before.

At least there is a C-L-C filter before all regulators.What value on the choke should i use after?The one before is 4,7uH.

put snubber on the power supply diode, or better change to fast "soft"recovery type

I´will do that.Thanks.
The e-lyts looks to be the cheapest type(non ESR)If I change them for
ESR types will it affect the PSU negaive in any way?(Switching PSU)
And just to be sure,if anyone is familiar with the schema,is it, "+5v_NSW" that is for the digital part?
The one marked +5v is probably for the rest,motors and stuff,I think.


The LM2940
seems pretty good on ripplerejektion,havent gotten any answer from texas yet.

[Hartono]

"The e-lyts looks to be the cheapest type(non ESR)"

I guess you mean High ESR


"If I change them for
ESR types will it affect the PSU negaive in any way?(Switching PSU)"

and you would change the with low ESR types.

sometimes this can cause oscillation (bad) in switching PSU

Edit : if you want to change it, change the one after the choke, larger value is ok



Hartono

[Ryssen]

Yes i meant high ESR,sorry..
Ok,so only them after then.

[anatech]

Hi Ryssen, Hartono,
Often, I have found the regular diodes to be the best. Don't always rush to replace them.

-Chris